báo cáo khoa học: "Talk, trust and time: a longitudinal study evaluating knowledge translation and exchange processes for research on violence against women" potx

RESEARCH Open Access Talk, trust and time: a longitudinal study evaluating knowledge translation and exchange processes for research on violence against women C Nadine Wathen 1* , Shannon L Sibbald 1,2 , Susan M Jack 3 and Harriet L MacMillan 4 Abstract Background: Violence against women (VAW) is a major public health problem. Translation of VAW research to policy and practice is an area that remains understudied, but provides the opportunity to examine knowledge translation and exchange (KTE) processes in a complex, multi-stakeholder context. In a series of studies including two randomized tri als, the McMaster University VAW Research Program studied one key research gap: evidence about the effectiveness of screening women for exposure to intimate partner violence. This project developed and evaluated KTE strategies to share research findings with policymakers, health and community service providers, and women’s advocates. Methods: A longitudinal cross-sectional design, applying concurrent mixed data collection methods (surveys, interviews, and focus groups), was used to evaluate the utility of specific KTE strategies, including a series of workshops and a day-long Family Violence Knowledge Exchange Forum, on research sharing, uptake, and use. Results: Participants valued the opportunity to meet with researchers, provide feedback on key messages, and make personal connections with other stakeholders. A number of factors specific to the knowledge itself, stakeholders’ contexts, and the nature of the knowledge gap being addressed influenced the uptake, sharing, and use of the research. The types of knowledge use changed across time, and were specifically related to both the types of decisions being made, and to stage of decision making; most reported use was conceptual or symbolic, with few examples of instrumental use. Participants did report actively sharing the research findings with their own networks. Further examination of these second-order knowledge-sharing processes is required, including development of ap propriate methods and measures for its assessment. Some participants reported that they would not use the research evidence in their decision making when it contradicted professional experiences, while others used it to support apparently contradictory positions. The online wiki-based ‘community of interest’ requested by participants was not used. Conclusions: Mobilizing knowledge in the area of VAW practice and policy is complex and resource-intensive, and must acknowledge and respect the values of identified knowledge users, while balancing the objectivity of the research and researchers. This paper provides important lessons learned about these processes, including attending to the potential unintended consequences of knowledge sharing. * Correspondence: nwathen@uwo.ca 1 Faculty of Information and Media Studies, The University of Western Ontario, London ON Canada Full list of author information is available at the end of the article Wathen et al. Implementation Science 2011, 6:102 http://www.implementationscience.com/content/6/1/102 Implementation Science © 2011 Wathen et al; licensee BioMed Central Ltd. This is an O pen Ac cess article distributed under the terms o f th e Creative Co mmons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background Data on the prevalence [1-3], consequences [4-7], and costs [8,9] of intimate partner violence (IPV) against women attest to its persistent and devastating impact on the lives of women, their children, and society. It has been almost 20 years since IPV was declared to be a major public health problem [10], yet many gaps remain regarding effective approaches to detecting and respond- ing to it [11-14], which have led to debates and conflict- ing advice to health and social service providers and policy decision makers [15]. McMaster Violence Against Women (VAW) Research Program and Knowledge Translation and Exchange (KTE) Project In 2003, the Ontario Women’s Health Council (OWHC), an advisory body to the Ontario Minister of Health and Long-Term Care, funded the VAW Research Program at McMaster University. The re search program had as its primary goal answering the question: does routine screening for intimate partner violence against women presenting to healthcare settings reduce violence and improve life quality for women? The program was conducted in three phases (Figure 1), with multiple qua- litative, quantitative, and mixed-methods projects designed to answer specific questions that required evi- dence in order to develop the main study, a randomized controlled trial (RCT) of the effectiveness of screening including 18 months of follow-up. In 2006, a group of researchers from the VAW Research Program, in part- nership with policy analysts from the OWHC, were funded to begin to identify and develop the main mes- sages arising from the completed and ongoing projects. In 2008, we received new funding for additional KTE activities focussed on the results of the screening effec- tiveness trial that were published in 2009 [16]. Approach to KTE The KTE activities described in this study were guided by the interactio n model of knowledge translation [17,18], and assumed that effective KTE would require initiating and assessing ‘various disorderly interactions occurring between researchers and users’ [18] and understanding that researchers and knowledge users (broadly defined) are ‘two communities’ [18], or in the case of our identified stakeholder groups, multiple com- munities. The McMaster VAW Research Program uti- lized an integrated knowledge translation approach [19] with knowledge users representing clinical practice, community service, and public policy decision-making constituencies involved from the outset as members of the research team. These partners, in addition to helping to shape the design of the research studies, were key resources when planning and implementing the KTE strategies described below. The interaction model also stresses the development and evolution of ‘relationships between researchers and users at different stages of knowledge production, disse- mination and utilization’ [18], and assumes that more numerous and intensive interactions between research- ers and users will lead to greater potential for use of the knowledge; this rationale underpinned our approach to multiple contacts (frequency and type) over time, both with organizations and individuals. Further, the factors that mediate knowledge utilizatio n include, according to Oh and Rich [17]: characteristics of the information, organizational characteristics, motivations and attitudes of the knowledge users, and the nature of the knowledge gap/problem to be addressed. The contextualization of research messages and KTE strategies to take into account the second and fourth factors above was a key priority in our KTE processes [20-22]. Finally, to map our findings across key stages of KTE processes as generally articulated in the literature [21,23], our questions were asked, and results are pre- sented, according to the fo llowing: dissemination and uptake; sharing and use; and impact. Within the ‘sharing and use’ domain, recognizing that ‘knowledge utilization’ is a multi-faceted phenomenon, we frame the concept in three ways consistent with the KTE literature [18,24,25]: conceptual/enlightening use (i.e., ‘to provide better understanding or insight about an issue’); symbolic/ selective use (i.e., ‘to support or refute an existing belief, policy, process, or course or action’); or instrumental/ direct use (i.e., ‘to propose a new policy, process or course of action’). The overall goal of our KTE project, therefore, was to ensure that results arising from the research were Figure 1 McMaster VAW Research Program. A schematic of the research program and projects from with the research evidence for the KTE project was drawn. VAW: violence against women; RCT: randomized controlled trial; PHN: public health nurse; KTE: knowledge translation and exchange. Wathen et al. Implementation Science 2011, 6:102 http://www.implementationscience.com/content/6/1/102 Page 2 of 15 identified early, developed appropriately, and shared with key audiences, including policy decision makers, healthcare practitioners, community service providers, and advocates. In this paper, we report on the four-year study that identified and documented how stakeholders received,engagedwith,andused(ornot)theresearch knowledge shared with them via a series of KTE strate- gies. Our specific research questions were: How do reci- pients of research ev idence perceive the utility of specific KTE strategies in the area of violence against women? What fact ors, according to those recei ving research evidence, influence the uptake, sharing, and use of the new knowledge? And what kinds of use are made of research findings? We also reflect on the ‘lessons learned’ from this longitudinal project that might be applicable to KTE efforts and evaluation more broadly. Methods Design A longitudinal cross-sectional design, applying concur- rent mixed data collection methods [26,27], was used to describe and assess our KTE processes and their impact on the types of knowledge utilization described above. Phase 1 of the study examined the process for develop- ing initial research messages and sharing them with sta- keholders at an interim point in the research program. Phase 2 focused on uptake and use of the final results of the screening trial. Table 1 provides an overview of the KTE activities conducted from 2006 to 2009, and the data collection methods (to April 2010) used to eval- uate their impact. Phase 1: 2006 and 2007 Key message development The team reviewed VAW re search program reports, including results from eleven projects (Figure 1, Phase 1 and ‘Testing Trial’ [28]), to identify relevant findings. Key messages were identif ied using a structured, itera- tive process, including input from the research team and key polic y partners/funders, and presented using appro- priate formats [29] (see Additional File 1). Stakeholder workshops and evaluation In October 2006, we held half-day workshops in Lon- don, Toronto, and Ottawa, Ontario attended by 82 sta- keholders. Each began with a networking lunch, followed by research project presentations, key mes- sages, and preliminary synthesis, with time for discus- sion. Participants were then divided into groups, facilitated by a research team member, and discussed two questions: ‘what are the implications of these find- ings?’ and ‘what should happen next?’ followed by reporting-back and plenary discussion. Primary evaluation methods were:anevaluationsur- vey immediately post-workshop (n = 75); an online follow-up survey about three months post-workshop (n = 33); and in-depth telephone interviews about six months post-workshop with participants who had con- sented to follow-up (n = 20). The evaluation survey con- sisted of 10 structured questions asking about work setting, role and decision-making responsibilities, level of previous involvement with our research, and experi- ences during the workshop, including overall usefulness. The online follow-up survey asked similar questions to those above, and questions regarding influence/impact of using the research, as well as ongoing interaction between stakeholders and the research team. The follow-up telephone interviews used a descriptive qualitative approach [30] to further probe the evaluation survey results and to explore the impact that the work- shops had on subsequent decision making. We purpose- fully sampled from the three workshop sites at least two stakeholders from each of the following groups: public policy, healthcare providers (hospital and community- based), social se rvice provider s, and women’s advocates. The semi-structured interview guide asked about their experience of the workshop, whether they had shared or used the research (and if so, how and to what effect), or planned to do so. Interviews lasted about an hour and were audio-recorded (with permission). Development of an online community of interest Participants at all three workshops endorsed the idea of an interactive website, using ‘Web 2.0’ technologies, to allow ongoing interaction; we therefore developed the ‘online community of interest’ (http://www.VAWCom- munity.ca; link no longer active). Launched in March 2007, and using a wiki platform, the site included static documents and information (e.g., summaries, meeting notes and slides, et al.) and interactive areas, where users were invited to edit meeting notes to reflect their memory of the discussions, and edit key messages to make them more relevant or user-friendly. The online follow-up survey was linked to the site. Phase 2: 2008 and 2009 Key message development As with Phase 1, an initial step was to develop key mes- sages from the screening trial in the context of both our previous messages and the broader evidence-base (Addi- tional file 1). A particular challenge was the nature of the main results of the trial: for one primary outcome, recurrence of violence, the d ifference was not statisti- cally significant; for the second, quality of life, there was a small clinically non-significant difference that also became statistically non-significant following multiple imputation to account for data loss. The differences between all secondary outcomes were not statistically significant, with the exception of depressive symptoms, which showed the same pattern as quality of life. To Wathen et al. Implementation Science 2011, 6:102 http://www.implementationscience.com/content/6/1/102 Page 3 of 15 help enhance the relevance and clarity of the results, we held meetings with key Ontario policy stakehol ders (see details in Additional file 2);weusedtheirinputto develop the final key messages. Family violence knowledge exchange forum In January 2009, we hosted the Family Violence Knowl- edge Exchange Forum in Toronto, Ontario. This included, in a ddition to members of the research team, policy makers (federal and provincial), health and social service leaders, women’s and children’s advocates, and other family violence rese archers from across Canada. This day-long interactive event featured brief research presentations (oral and poster), plenary discussions, and 10 small group discussions that followed presentation of key messages [31]. Given feedback from the Phase 1 workshops, we minimized presentation of data, and maximized time for discussion; key messages were pre- sented as ‘actionable’ [32] messages in 10 minutes with minimal data, graphs, or research jargon. Forum evaluation In addition to audio- and video recording the session, field notes and post-meeting debriefing, we used the fol- lowing methods to evaluate the forum: Small group discussions (n = 10 groups) To understand the initial impact of the screening trial results, we captured stakeholders’ reactions to the key messages by a sking them to consider and discuss them immediately after they were presented (see discussion questions in Additional file 2). The discussion at eac h table was audio-recorded (those who did not want their comments recorded could pause the recorder while speak- ing). Based on feedback from the Phase 1 workshops, there was no formal group moderation; rather, the research team circulated to answer questions regarding the research. Table seating of 8 to 10 participants was pre- assigned to mix groups by sector, role, and geography. Evaluation survey (n = 38) Attendees were asked to complete an exit survey that used the same q uestions and format as the Phase 1 workshop evaluation. Follow-up survey (n = 21) Appr oximately six months after the forum, stakeholders who gave permission for follow-up were sent an email invi tation to complete an online survey. The survey had 18 structured questions similar in content to those described above. Table 1 Overview of knowledge translation and exchange (KTE) activities and evaluation strategies KTE Activity Description and Participants Evaluation Approach Phase 1 (2006 and 2007) Key message development (VAW Research Phase 1 Studies and Testing Trial) (Spring and Summer 2006) Core research team and policy partner/funder drafted key messages; the wider VAW research team reviewed them, and they were formatted for stakeholder audiences. Observation and journaling by core research team re: process Stakeholder workshops (October 2006) 82 stakeholders attended one of three half-day workshops in Toronto, Ottawa, or London Ontario. Workshop evaluation survey (Fall 2006) (n = 75) In-depth telephone interviews (Winter 2006/7) (n = 20) Follow-up online survey (Winter 2007) (n = 33) Online community of interest (launched Spring 2007) In response to request from stakeholders, created an online wiki- based site to continue interaction. Usage data Phase 2 (2008 and 2009) Key message development (Screening Trial) (late 2008 to early 2009) Core research team and policy partners/funders drafted key messages; wider VAW research team reviewed them, and they were formatted for stakeholder audiences, including media talking points. Observation and journaling by core research team re: process Family Violence Knowledge Exchange Forum (January 2009) Day-long meeting, in Toronto, of 76 stakeholders and 11 1 researchers from the McMaster VAW Research Program. Focus on high-level key messages and discussion of policy and practice implications. Forum evaluation survey (n = 38) Analysis of Forum small group transcripts (n = 10 groups) Participant follow-up survey (Summer 2009) (n = 21) Follow-up interviews (Fall 2009 to Winter 2010) (n = 12) Media (Summer 2009) Publication of screening trial in JAMA in August 2009 led to significant media interest Included questions about media exposure in follow-up interviews 1 Three members of the research team who attended were knowledge user partners with clinical leadership roles; VAW: violence against women; JAMA: Journal of the American Medical Association Wathen et al. Implementation Science 2011, 6:102 http://www.implementationscience.com/content/6/1/102 Page 4 of 15 Follow-up interviews (n = 12) Qualitative semi-structured interviews using the same methods described above for the Phase 1 follow-up interviews, and pr obing the same kinds of questions as the Phase 2 follow-up survey, were conducted 9 to 12 months following the forum. Additional KTE activities not directly assessed in this project are described in Additional File 2. Data Analysis Data from post-meeting and follow-up evaluation sur- veys were entered into Excel and/or SPSS, cleaned and checked by a research assistant, and descriptive statistics generated. Transcript data from meeting small groups and follow-up interviews were transcribed verbatim, cleaned, organized in NVivo © , and analysed using direc- ted content analysis [33], with each coder using a list of predetermined codes based on the concepts explored in the interview or group. Codes were then collapsed into primary categories. To ensure trustworthiness of the data, each transcript was independently reviewed, and key themes identified, by two study investigators, with rev iew of synthesized results by additional coll aborators who had attended the workshops and forum. Concur- rent triangulation of t he results [27] within each phase allowed us to integrate the qualitative and quantitative data for more complete interpretation of participants’ experiences and perspectives, as well as using emerging findings from Phase 1 to inform the development of the KTE strategies and evaluation methods used in Phase 2. Ethical considerations Phase 1 of the study was reviewed and provided a waiver (i.e., deemed to pose no potential risk to consent- ing participants) by the McMaster University Faculty of Health Sciences- Hamilton Health Sciences Research Ethics Board (REB). Phase 2 was approved by the Uni- versity of Western Ontario Non-Medical REB (protocol #15789S). Results Participant Characteristics Given the nature of the data collection methods and ethical requirements regarding participant anonymity, each data collection point represents a separate sample– i.e., this is not a cohort of individuals followed across time, but rather individuals who self-selected participa- tion at these various points in the study; 190 stake- holders were invited to the 2006 workshops and 82 attended; 217 were invited to the forum, and 76 attended; 139 stakeholders (34%) were invited to both events, and 15 (8.9%) attended both. Thus, while there was minimal overlap between the samples of respon- dents to our data collection approaches, there was certainly growing awareness of the work among the overall targeted group of stakeholders (individuals and organizations) who received invitations and interacted with the research office re: RSVPs and other meeting logistics. Tables 2 and 3 describe the types of work- places (Table 2) and decision-making roles (Table 3) reported by respondents in Phases 1 and 2. A wide range of settings and roles were represented, with 56% of participants reporting having multiple decision-mak- ing roles, and a significant number reporting an overlap between clinical/service delivery and planning/adminis- trative roles. Additional file 2 provides an overview of the samples participating at each stage of data collec- tion, and specific sub-sample sizes are specified in the Tables. To understand the relationships between the research team and stakeholders, we asked about their previous involvement with the VAW research program. In gen- eral, most respondents indicated low involvement, including receiving information on study findings through formal (Phase 1: 27%; Phase 2: 22%) and infor- mal (32% and 25%, respectively) processes, or simply being ‘aware of the VAW research program but not much else’ (32% and 17%, respectively) (the decrease in this response over time may reflect respondents’ expo- sure to earlier KTE efforts); 1 5% (Phase 1) and 22% (Phase 2) were ‘not aware of the program until invited’ to the event. Thus, the stakeholders to whom we spoke had varying, but generally not well developed, familiarity with the research program and its emerging findings. Knowledge uptake, sharing, use, and impact: Key findings In order to examine key aspects of our KTE processes and the uptake and use of findings by stakeholders, the results of the study are presented across the study phases and according to the KTE activities and stake- holder reactions to them, while attempting to describe how the research knowledge was heard, shared, and used, and what, if any, early impact it may have had. Because the quantitative survey questions w ere highly complementary with the qualitative interview questions, we present related data together–that is, proportions of participants responding to survey questions are pro- vided, and supporting quotes from write-in comments and interview transcripts are used to highlight and eluci- date key findings regarding the KTE stages. Analysis of the content of the reactions, and their implications for VAW policy and practice, are beyond the scope of this paper. Knowledge dissemination and uptake The focus of this section is to highlight participants’ perceptions of our KTE processes, and identify which strategies were effective, and which were not. Wathen et al. Implementation Science 2011, 6:102 http://www.implementationscience.com/content/6/1/102 Page 5 of 15 Across both phases of this work, it was clear that partici- pants placed significant value on the opportunity to attend in-person meetings with researchers, and with other stake- holders, and to think a bout a nd discuss research and its potential impact on decision-making in ways not usual to their daily work. Especially helpful, said participants, was the opportunity to meet and discuss the research with the researchers (80% of workshop and 92% of foru m parti ci- pants reported that they had the opportunity to meet and discuss the research with one or more of the VAW researchers and 97% and 100%, respectiv ely, found this valuable), which helped them to assess the credibility and quality of the research. As one commented: ’It’s hard to know the quality of [research] because we all skim these days. We all rely on just skimming through things and saying, ‘Okay, is this something I should read in more depth?’ And when you know the researchers by reputation, then I know that anything with their name o n it is going to be worth reading.’ [workshop follow-up interview, P06] This ability to make a personal connection with a researcher enhanced awareness of the research and put it on their ‘radar screen,’ increasing the likelihood that future communiqués from the team would stand out. It was highlighted that face-to-face meetings are an impor- tant step in building relationships: ’ having the ability to [not just read] a paper but to hear from the researchers themselves and have the time and the luxury to digest and distil the infor- mation I think just keeps this research on top of the pile as opposed to getting lost in the shuffle of the many pieces of research that cross our desks.’ [forum follow-up interview, P08] In terms of the workshop and forum as information- Table 2 Workplace types reported by participants (Phase 1 workshop evaluation and online follow-up survey; Phase 2 forum evaluation) Workplace type 1 Phase 1: Workshop Evaluation (n = 75) % (n) Phase 1: Follow-Up Survey (n = 33) % (n) Phase 2: Forum Evaluation (n = 38) % (n) Community based service organization (e.g., Shelter) 24% (18) 9% (3) 8% (3) Advocacy group 9% (7) 0% (0) 3% (1) Acute or primary healthcare service organization 32% (24) 30% (10) 13% (5) Public health unit or agency 23% (17) 21% (7) 13% (5) Government department (provincial, federal, municipal) 17.3% (13) 24% (8) 16% (6) University department and/or research centre 7% (5) 0% (0) 34% (13) Other (write-in) 5% (4) 2 15% (5) 3 13% (5) 4 1 respondents could indicate more than one type; 2 legal, police, professional association, grassroots; 3 university student, professional association, municipal government, police, information centre; 4 minister, lawyer, regulatory body, provincially funded support and funding agencies, police Table 3 Types of decision-making roles (Phase 1 workshop evaluation and online follow-up survey; Phase 2 forum evaluation and online follow-up survey) Decision-making role Phase 1: Workshop Evaluation 1 (n = 74, 1 missing) % (n) Phase 1: Follow-Up Survey 2 (n = 33) % (n) Phase 2: Forum Evaluation 1 (n = 35, 3 missing) % (n) Phase 2: Follow-Up Survey 2 (n = 16, 5 missing) % (n) Clinical care/service delivery decisions 41% (30) 12% (4) 23% (8) 6.3%(1) Planning/programming decisions 51% (38) 18% (6) 31% (11) 31.3%(5) Administrative decisions 41% (30) 27% (9) 9% (3) 0% (0) Public policy decisions 19% (14) 15% (5) 17% (6) 18.8%(3) Research decisions 4% (3) 9% (3) 14% (5) 12.5%(2) Advocacy decisions 32% (24) 3% (1) 14% (5) N/A Other (write-in) 20% (15) 3 15% (5) 4 26% (9) 5 31.3%(5) 6 1 respondents could indicate more than one role; 2 respondents were asked to indic ate one decision role only 3 including, in order of frequency: education and training (of other professionals, or public awareness/outreach); other policy work; funding to implement public policy; 4 management and mixed roles re: service, advocacy, research and planning; 5 including project management, regulatory, policy recommendations, training, community presentations, policing, funding; 6 including education/curriculum, funding, communi cation a nd knowledge translation. Wathen et al. Implementation Science 2011, 6:102 http://www.implementationscience.com/content/6/1/102 Page 6 of 15 sharing venues, there was certainly acknowledgement that these types of events increased their understanding of the complexity of the research process, including clarification of study findings and limitations; however, there was still a feeling from participants of wanting more–more clarity in what the data were saying, and more direction on what the data means for future prac- tice: ’I think I was hoping to get more specific detail on some of the studies. More o n identifying [and] responding to intimate partner violence in healthcare settings.’ [forum follow-up interview, P05] Participants also appreciated pre-circulated materials and handouts, and, especially at the forum, having the key messages ‘well explained and clearly presented’: ’ the research data had been boiled down to key messages and I know how dif ficult that was for the researchers. Really was much more impactful than a whole series of conclusions and you get lost in the information.’ [forum follow-up interview, P03] There was also some concern of information overload; however, this was balanced out with the appreciation for getting the larger picture, and making it relevant to a variety of stakeholders. These experiences were slightly different between the workshops, which presented much more detail regarding a series of individual projects, and the forum, which, based on feedback from the work- shops, presented key messages concisely and clearly. Another important experience for participants was the opportunity to provide feedback on the ke y messages (94% in the workshops and 98% in the forum reported this, and 98% of both groups found this valuable). In the workshop follow-up interviews, participants identified that providing feedback on a study still in progress was a novel and positive experience, especially for frontline staff from community-ba sed services. As one workshop participant said: ’[The workshop used a] truly collaborative approach [with] respect for the input of the frontline. Often research is presented as a done dea l, and frontline advocates, who I would say are the experts on the subject matter, are just treated as the consumers of the information versus the creators or holders of the information. I thought the [workshop] process was reallyrespectfulandthatitworkedreallywell.’ [workshop follow-up interview, P07] Several participants highlighted the necessity of a ‘common language’ and a common space for these sorts of discussions–and this type of forum was a good step in that direction, but that more still needed to be done. When asked how responsive the project team was to their ideas and suggestions, most found us very or somewhat responsive (workshops: 80%; forum: 75%), while the rest indicated it was ‘toosoontotell’ or ‘not applicable.’ Individuals interviewed overwhelmingly described that the research team was genuin e and respectfully listened to the different pe rspectives of VAW shared by participants. Another significant benefit highlighted by participants was the opportunity to network with peers from across sectors and the multiple chances to engage in both indi- vidual and large and small group d iscussions: partici- pants reported that they had an opportunity to meet other stakeholders (workshop: 94%; forum: 95%), which they found very valuable (99% in both samples); the opportunity to network over a meal was also appre- ciated. For some participants, the workshop provided a venue to share information about their organization and the services it offers. A small number of workshop participants from the same group commented verbally to a member of the research team that those discussions were not well- facilitated, that the facilitation interfered with genuine discussion, or there was a single individual who domi- nated the conversation (and was not well-handled by the facilitator). Based on this feedback (which was pro- vided informally and was not reflected in the w ritten evaluations), formal facilitation was not used during the forum, and there were no expressed concerns regarding those discussions. Participants from both events liked the group format, espec ially mixing the stakeholders, as expressed by this person: ’ to be at a table with folks that were coming from different perspectives and having that conversa tion on how these messages were being interpreted by thosedifferentperspectivescertainlygavemesome food for thought in terms of how do you communi- cate these messages to people who really need to hear them. When you know they may [be] hearing different things than wha t you are trying to say.’ [forum follow-up interview, P08] When asked how valuable, overall, the events were for them and their work, the majority of participants indi- cated very or somewhat valuable (workshop 80%, forum: 89%); with the rest indicating it was ‘too soon to tell’ (workshop: 19%, forum: 11%). We also asked partici- pants if they would like to stay connected with our KTE processes, and nearly all those who responded said they would(workshop:97%;forum:95%).Whiletheyindi- cated a range of preferences for ongoing communication Wathen et al. Implementation Science 2011, 6:102 http://www.implementationscience.com/content/6/1/102 Page 7 of 15 with the research team, of note is the preference for electronic approaches, with over 75% preferring being sent electronic summaries of findings and/or links to the program website when new material is posted. Despite the significant enthusiasm among workshop participants for the wiki-based ‘online community of interest,’ beyond an initial visit for it s launch in March 2007 and completion of the follow-up survey, and despite encouraging reminders, the wiki was never used, and was eventually removed late in 2007. Knowledge sharing and use This section presents data from the workshop and forum follow-up surveys and interviews specific to whether and how people shared what they had heard, and whether and how they had used the research find- ings. Table 4 provides an overview of the quantitative data from the two surveys, which is discussed below in light of what participants said during the interviews. Sharing research knowledge As indicated in Table 4, in the three to six months fol- lowing each event, the majority of participants indicated that they had shared the research with people in their organization and/or with external colleagues (workshop: 88%; forum: 79%); the information was shared verbally, electronically, through document-sharing and via reports and presentations: ’I did a lunch-and-learn with my colleagues about the research presented at this forum. I also pre- sented the knowledge and my reflections on the event to our management team. ’ (forum follow-up survey, write-in). Of interest is the type of sharing activity reported by those who attended the forum, all of whom reported more recipients of and app roaches to sharing info rma- tion, including internally, and also more broadly (93%) beyond ‘colleagues.’ Those who reported not sharing the information indicated that the primary reason for this was ‘lack of opportunity.’ Using research knowledge The bottom part of Table 4 indicates that while there was some reported ‘use’ of the findings at the three- to six-month post-event point, this occurred much less oftenthanthe‘sharing’ of knowledge, and was more consistent between the two phases, perhaps indicating that finding ways to actually integrate research evidence into decisions–especially after a relatively short period of time–is a much more complex process than simply ‘passing it on.’ In terms of use (and keeping in mind the small sub-samples who indicated use of any kind) across both phases (10 and 7 people, respectively), it was more common for the knowledge to be used symbolically and/or conceptually than instrumentally. The follow-up interviews (at approximately 12 months post-event) helped shed some light on these processes. Reflections from participants indicated that in some cases the r esearch findings increased their understand- ing (conceptual use) of issues related to VAW, and that when this was the case, findings were more likely to be used to reinforce or support current policies or pro- grams within their organization (symbolic use). F or example, in the 12-month forum follow-up interviews, ten participants used the information conceptually as background or context for other work they were doing. In this way, the information heard at the forum pro- vided a new lens, and an opportunity to further consider their current practices: ’Well I think you take it more personally. I think you try to apply it to your everyday kno wledge and your experience when you are front line.’ [forum - small group 9] Some participants used the research findings more instrumentally, for example incorporating it into in- house employee training, or into a report, or to update cli nical protocols or guidel ines. Forum participants who had used the information cited their attendance at this event as a major facilitator to using this knowledge. However, we also learned that a number of participants would choose not to use evidence from the research pro- gram in thei r decision-making when it contradicted their personal experiences. These participants expressed dis- comfort with specific key messages (e.g., that screening is ineffective, or that pregnancy was not a risk indicator fo r current abuse). One workshop participant, even several years prior to completion of the screening trial, stated: ’Fromourexperiencewehavealreadyproven,or believe that we have proven that they [protocols for universal screening] have been incredibly effective and we will continue to have that policy and proce- dure in place So I would say it [the research evi- dence] has l ittle or no impact ’ [workshop follow- up interview, P12] And, during discussion of the actual trial results at the forum, another said: ’Well, we thought it would be unfortunate if the research was used to discredit the value of universal screening because intuitively we felt that universal screening made some sense even though the research doesn’tshowthatit’s probably worth the Wathen et al. Implementation Science 2011, 6:102 http://www.implementationscience.com/content/6/1/102 Page 8 of 15 resources and the effort to do it. The benefits aren’t worth that. But so we felt that perhaps it was prema- ture to say that it wasn’t.’ [forum small group 5] With regard to evidence around pregnancy and risk for violence: ’You know I’ ve heard stories around this from women [that when they become pregnant, the abuse starts], so anecdotally I know that it’strue.When the McMaster study said that preg nancy was not a [risk] indicator, I said, and was supported by other VAW people, ‘That doesn’t fit with our experience.’’ [workshop follow-up interview, P16] Impact It is well-acknowledged in the KTE literature that the most difficult thing to assess is the actual impact of new knowledge on specific p olicies or practices, or, ulti- mately, on health-related outcomes. We therefore exam- ined t he notion of ‘impact’ in terms o f what our participants reported with re spect to both the effect of them sharing the new knowledge with others, their own assessment of what happened when they used the research findings, and finally, the impact of the KTE processes themselves on respondents’ decision-making. Impact of sharing–how do others respond to the knowledge? The first aspect of ‘impact’ relates to how others reacted when participants share d the research findings with them. In general terms, we wanted to know whether the reactions were positive or negative (or neither), and what people might be planning to do with th is new knowledge. Of the 22 respondents who shared the knowledge from the workshops, 43% indicated a positive reaction, 29% a negative reacti on, and 24% were unsu re Table 4 Sharing and use of research results - follow-up surveys (Phases 1 and 2) Question Phase 1 Follow-Up Survey (3 to 6 months post-event) (n = 25, 8 missing) Phase 2 Follow-Up Survey (6 to 8 months post-event) (n = 21) Shared the research knowledge from the event? YES = 88% (22) NO = 12% (3) YES = 79% (15 of 19 who responded) NO = 21% (4) For those who responded YES n = 22 n = 15 Shared with (all that apply): Internal colleagues 42% (10) 100% (15) External colleagues 48% (11) 47% (7) Others 14% (3) 93% (14) How shared (all that apply): Verbally 46% (11) 93% (14) By email 4% (1) 13% (2) Sent documents 8% (2) 47% (7) Other 1 25% (6) 53% (8) Response to sharing Positively 43% (9) 27% (4) Negatively 29% (6) 13% (2) Mixed/Neutral 0% (0) 53% (8) Can’t tell/other 29% (6) 13% (6) Used the research knowledge from the event? YES = 40% (10) NO = 60% (15) YES = 37% (7) NO = 63% (12) Missing = 2 For those who responded YES n = 10 n = 7 How used (all that apply): (1 missing) (2 missing) Conceptual 50% (5) 80% (4) Symbolic 40% (4) 80% (4) Instrumental 0 40% (2) Have others used the research knowledge from the event? YES = 12% (3) NO/Don’t Know = 88% (22) YES = 26% (5) 2 NO/Don’t Know = 74% (14) (2 missing) 1 including: providing a link to the Research Program (n = 6), formal presentations or reports (n = 1) and informal discussions (n = 7); 2 In the Forum follow-up survey we asked how others had used the knowledge, of the three responses, two indicated conceptual use, and one indicated instrumental use. Wathen et al. Implementation Science 2011, 6:102 http://www.implementationscience.com/content/6/1/102 Page 9 of 15 of the reaction; write-in comments on the workshop fol- low-up survey, including one ‘other’ remark, indicated a range of reactions to the research, from colleagues pre- ferring to wait for the final results (of the screening trial) to disappointment in hearing that abused women would prefer computer-based screening to speaking to a healthcare provider. Whe n asked the same question, 27% of the 15 respondents from the forum indicated a positive reaction, 13% a negative reaction, and 53% a mixed reaction. Write-in comments, indicating the diversity of reactions, included: ’The findings somewhat discouraged some people, as the y had seen screening as the answer to addressing this gap.’ ’The screening issue continues to be a hotly debated subject and, while we are excited about the direction of the findings, those who are committed to screen- ing continue to dispute the direction to expand beyond screening.’ Impact of knowledge use–what happened? We asked in slightly different ways in the two follow-up surveys the question of ‘what happened’ with regard to using the research findings.Fortheworkshops,the focus was on the impact of use on a 5-point scale from very negative to very positive, or the option ‘too soon to tell.’ In that survey, of the 10 people who indicated they had used the research, six indicated the impact of this use was very positive or positive, one indicated it was negative, and three said it was too soon to tell. In the forum follow-up survey and interviews, we asked ‘what happened as a result of using the research findings on woman abuse screening?’ and gave some specific response options, with respondents asked to check ‘all that apply’ and also comment on whether there was impact in their own work, and/or in the work of others. Only four people answered this question. Of those that did respond, the impact included actual or proposed/planned change to a policy, process, or course of action, and new points of discussion about these. None of these people expected ‘nothing’ to happen as a result of using the research knowledge, and when we asked participants to rank the impact the information has had on their work on a scale of 1 to 5 (5 is high), most (90%) felt quite positive about the impact, saying it hadanimpactofbetween3and5.However,thediffi- culty in assessing ‘impact’ was reflected by this intervi e- wee: ’Well that’s a really hard question to answer because on the one hand absolutely no impact because we were alread y [decided against screening] sup porting thatandtheyweresupportingourwork,sonone. And then at the same time it’ sabsolutelyhigh because it affirmed in a kind of more objective way whatweweredoing.SoexternallyIthinkit’sa five [ranking]; internal for our own work, not so much.’ [forum follow-up interview, P11] In the forum follow-up survey, we asked if they planned to use the results in the future. Of the 18 who responded, 61% said yes and one person said no; the remaining six indicated it was too soon to tell. Write-in responses to the forum follow-up survey for this ‘poten- tial use’ question included those who intended very spe- cific uses: ‘we plan to use the findings to develop formal woman abuse policy at our hospital as well as to direct provincial policy’; ‘we are establishing a core public health program on the prevention of violence and abuse this issue will be discussed at our first working group meeting ’; and the potential conceptual impact described by this participant: ’ may be useful in exploring why screening (or screening + brief intervention) should be seen as a prelude to treatment. I am interested in factors asso- ciated with treatment engagement and findings from this study may provide background support for the need to consider screening as a first-step in engaging people in treatment.’ [forum follow-up survey, write- in] In contrast, some were quite clear that the results contradicted their practices, and hence would not be used, or would be used selectively to support current approaches: ’the research indicated that universal screening for IPV does not cause harm; therefore, I will be using this research to continue to advocate for universal screening of IPV,’ and: ‘our students are currently taught to screen for abuse and this would create a mixed message.’ Impact of KTE strategies and process Finally, we wanted to understand the impact that parti- cipating in our KTE activities had on participants. When we asked whether they thought that being at one of our events would influence their decision making, among the workshop participants 42% said yes, 3% said no, and 49% said it was too soon to tell (four people either gave multiple responses or did not answer). These participants were also asked whether attending the workshop had influenced their decision making: 35% said ‘yes’ and 65% said ‘no’; with regard to the overall Wathen et al. Implementation Science 2011, 6:102 http://www.implementationscience.com/content/6/1/102 Page 10 of 15 [...]... theories and Page 14 of 15 methodologies that can assess and explain knowledge mobilization’ as a construct related to, but distinct from, current knowledge translation approaches are required Additional material Additional file 1: Key Message Development for the VAW Research Program and Messages Presented at January 2009 Family Violence Knowledge Exchange Forum and Additional Background Information About... Disease Control and Prevention, Atlanta (GA); 2003 Bonomi AE, Anderson ML, Rivara FP, Thompson RS: Health care utilization and costs associated with physical and nonphysical-only intimate partner violence Health Serv Res 2009, 44(3):1052-1067 American Medical Association, Council on Scientific Affairs: Violence against women: relevance for medical practitioners JAMA 1992, 267(23):3184-3189 Wathen CN, MacMillan... McMaster VAW Research Program Additional file 2: Additional details regarding study processes, data collection tools, methods and samples Acknowledgements Phase 1 of the VAW-KTE project was funded by a Canadian Institutes of Health Research (CIHR) Knowledge to Action Grant Phase 2 was funded by a Social Science and Humanities Research Council of Canada Presidents Initiative Grant on Capturing the Impacts... social and behavioral research Edited by: Tashakkori A, Teddlie C Thousand Oaks CA: Sage; 2003: Tashakkori A, Teddlie C: Mixed methodology: combining qualitative and quantitative approaches Thousand Oaks CA: Sage; 1998 MacMillan HL, Wathen CN, Jamieson E, Boyle M, McNutt LA, Worster A, et al: A randomized trial of approaches to screening for intimate partner violence in health care settings JAMA 2006,... mental disorders: a meta-analysis J Family Violence 1999, 14(2):99-132 Sharps PW, Laughon K, Giangrande SK: Intimate partner violence and the childbearing year: maternal and infant health consequences Trauma Violence Abuse 2007, 8(2):105-16 Centers for Disease Control and Prevention, National Center for Injury Prevention and Control: Costs of IPV against women in the United States Centers for Disease... and Media Studies, The University of Western Ontario, London ON Canada 2Faculty of Health Sciences, The University of Western Ontario, London ON Canada 3School of Nursing, McMaster University, Hamilton, Ontario, Canada 4Departments of Psychiatry and Behavioural Neurosciences, and of Pediatrics, Offord Centre for Child Studies, McMaster University, Hamilton, Ontario, Canada 1 Authors’ contributions CNW... Funded Research; the January 2009 Family Violence Knowledge Exchange forum was funded by a CIHR Meetings, Planning and Dissemination Grant–End of Grant Knowledge Translation Supplement The McMaster Violence Against Women Research Program was funded by Echo: Improving Women’s Health in Ontario (formerly the Ontario Women’s Health Council Nadine Wathen is supported by a CIHR Institute for Gender and Health... Investigator Award in Women’s Health Susan Jack is supported by a CIHR Institute of Human Development, Child and Youth Health, Reproduction and Child Health New Investigator Award Shannon Sibbald is supported by a Canadian Health Services Research Foundation post-doctoral fellowship Harriet MacMillan is supported by the David R (Dan) Offord Chair in Child Studies Author details Faculty of Information and. .. Hanna SE, Ciliska D, Manske S, Cameron R, Mercer SL, O’Mara L, Decorby K, Robeson P: A randomized controlled trial evaluating the impact of knowledge translation and exchange strategies Implement Sci 2009, 4(1):61 44 Estabrooks CA, Thompson DS, Lovely JJ, Hofmeyer A: A guide to knowledge translation theory J Contin Educ Health Prof 2006, 26(1):25-36 45 Sales AE, Estabrooks CA, Valente TW: The impact... policies, as well as the timing of decisions that may (or may not) incorporate research findings Our findings in this area have implications for how we think about knowledge translation more broadly In fact, in an area such as violence against women, the ‘evidence-based medicine’ framework may well be inappropriate for knowledge designed to inform not only health services, but also broader community and . RESEARCH Open Access Talk, trust and time: a longitudinal study evaluating knowledge translation and exchange processes for research on violence against women C Nadine Wathen 1* , Shannon. 13:188-190. doi:10.1186/1748-5908-6-102 Cite this article as: Wathen et al.: Talk, trust and time: a longitudinal study evaluating knowledge translation and exchange processes for research on violence against women. Implementation Science. States Centers for Disease Control and Prevention, Atlanta (GA); 2003. 9. Bonomi AE, Anderson ML, Rivara FP, Thompson RS: Health care utilization and costs associated with physical and nonphysical-only